B. Sarmento et al. / Colloids and Surfaces B: Biointerfaces 53 (2006) 193–202
Taking into account these considerations, dextran sulfate(DS), a biodegradable and biocompatible branched negativelycharged polyion with approximately 2.3 sulfate groups per glu-cosyl residue, was selected to formulate DS/chitosan nanoparti-cles. Chitosan and DS have been formulated as polyelectrolytecomplex gels with swelling characteristicsand controlled
release propertiesand have been extensively used as multi-
layer ﬁlms for cell culture.Additionally, they are intrin-
sically hydrophilic, which contributes for longer circulationtimes
and allows the encapsulation of water-solubleproteins.In the present work insulin was chosen as the model protein.Itisawellknown51aminoacidsprotein,producednowadaysbyDNA recombination techniques and used subcutaneously in thetreatment of diabetes
. Although several attempts havebeen developed regarding alternative routes of insulin adminis-tration[19–21],the oral approach remains the most attractive
due to convenience and high patient compliance. However, thebioavailability of insulin after oral administration is normallylow, due to its low stability in the gastrointestinal tract (GI),lowpartitioncoefﬁcientandthephysicalbarrieroftheintestinalepithelium.Oneapproachtoimprovethegastrointestinaluptakeoflowmolecularweightproteinsistobindthemtocolloidalsys-temslikenanoparticles,protectingthemfromdegradationinthegastrointestinal tract and promoting the transport into systemiccirculation[22,23].Sincethesizeofnanoparticlesandproteinstabilityareessen-tial for an efﬁcient pharmacologic effect of insulin, the mainobjective of this paper was to establish a nanoparticulate sys-tem produced by polyelectrolyte complexation between DS andchitosan, and evaluate their potential as insulin carriers. Thephysical and morphological properties of nanoparticles wereinvestigated in accordance with formulation parameters and thereleaseproﬁleofinsulinwasalsodeterminedregardingitspoten-tial for oral delivery.
2. Materials and methods
Low molecular weight (MW) chitosan (
50kDa) and lowMW dextran sulfate (8kDa) were purchased from Sigma (Por-tugal). High MW dextran sulfate (500kDa) was obtained fromPKC
(Denmark). Dextran sulfate stock solutions were pre-pared in deionized water (Milli-Q
) by overnight magneticstirring and chitosan solutions were prepared by dissolutionof chitosan in deionized water containing 1% (v/v) acetic acidfollowed by ﬁltration through a paper ﬁlter Millipore #2 andstored at 4
C. Human zinc-insulin crystal (Lot RS0325, 7.0mglyophilized human biosynthetic insulin per vial) was a gift fromLilly Portugal.
2.2. Preparation of nanoparticles
Nanoparticle complexation between DS and chitosan wasperformed employing aqueous solutions of oppositely chargedpolymers in a ﬁnal volume of 120mL. Unless otherwise men-tioned, complexes were obtained after dropwise addition of chitosan solution at pH 5.0 to DS solution at pH 3.2 undermagnetic stirring followed by additional mixing for 15min at600rpm to ﬁnal concentrations of 0.15% DS and 0.10% chi-tosan (DS/chitosan mass ratio 1.5:1). For insulin association,7mg of the protein was previously dissolved in the DS solutionbefore chitosan complexation. These conditions were obtainedafter preliminary studies that provided the best results takinginto account the desired mean particle size and insulin associa-tion efﬁciency.
2.3. Nanoparticle characterization
Measurements of particle size were performed by photoncorrelation spectroscopy (PCS) at 25
C with a detection angleof 90
and zeta potential by laser Doppler anemometry (LDA)using a Malvern Zetasizer and Particle Analyzer 5000 (MalvernInstruments) (
2.4. Stability of nanoparticles
The stability of nanoparticles was determined monitorizingtheir mean particle size and zeta potential over time. After pro-duction, nanoparticles were stored at 4
C in aqueous solutionand then the size and zeta potential were measured after 0, 7,14, 28, 42 and 60 days of shelf-life.
2.5. Isolation of particles
)for45min.Supernatantwereusedforinsulindeter-mination. Nanoparticles were kept at 4
C for further experi-ments.
2.6. Insulin association efﬁciency
Theamountofinsulinassociatedwiththeparticleswascalcu-lated by the difference between the total amount used to preparethe particles and the amount of insulin present in the aqueousphase after centrifugation. The association efﬁciency (AE) wasdetermined indirectly applying the following equation:AE
total amount of insulin
free insulin in supernatanttotal amount of insulin
2.7. In vitro release studies
Nanoparticleswereplacedinto20mLofUSPXXVIbuffers,namely hydrochloric acid pH 1.2, acetate pH 4.5, acetate pH 5.2and phosphate pH 6.8 and incubated at 37
C under magneticstirring. At pre-determined time intervals, samples were takenfor insulin determination and replaced by fresh medium.
2.8. Insulin determination
Insulin concentration was determined spectrophotometri-cally using the Coomassie PlusTM Bradford Assay (Pierce,